Author Topic: DIY automatic total cloud cover measurement  (Read 4499 times)

0 Members and 1 Guest are viewing this topic.

Offline Aussie Susan

  • Senior Member
  • **
  • Posts: 76
DIY automatic total cloud cover measurement
« on: November 08, 2016, 09:23:08 PM »
(Hope this is the right forum - apologies if not but I couldn't see a better one.)

Looking at some of the weather sites, they have a 'total cloud cover' entry for a site. I'm also aware that this is something that many (at least manual) weather stations report (in oktas).

A search of the web shows all sorts of ideas for detecting the presence of cloud, but many are related to astronomy and also to the area of sky near to where the telescope is pointing. I've also seen devices suggesting that the IR temperature of clear sky vs cloud can be used but again many of these seem to have a fairly limited view angle.

I'm looking for something that:
1) I can construct myself (with reasonably accessible parts)
2) covers the whole sky (or a major part of it)
3) can be used day and night (not sure about having the sun through a lens onto a CCD device but happy to be proven wrong)
4) can be connected to a bit of computing power (e.g. Raspberry Pi or microcontroller) for analysis of the sensor output

I have an engineering background (although I'm a project manager these days) and I'm not afraid of building things (mechanical or electrical/electronic)

Any ideas?

Susan

Offline Michaelpt

  • Member
  • *
  • Posts: 28
    • Weather-Above.com
Re: DIY automatic total cloud cover measurement
« Reply #1 on: December 17, 2016, 11:03:19 AM »
(Hope this is the right forum - apologies if not but I couldn't see a better one.)

Looking at some of the weather sites, they have a 'total cloud cover' entry for a site. I'm also aware that this is something that many (at least manual) weather stations report (in oktas).

A search of the web shows all sorts of ideas for detecting the presence of cloud, but many are related to astronomy and also to the area of sky near to where the telescope is pointing. I've also seen devices suggesting that the IR temperature of clear sky vs cloud can be used but again many of these seem to have a fairly limited view angle.

I'm looking for something that:
1) I can construct myself (with reasonably accessible parts)
2) covers the whole sky (or a major part of it)
3) can be used day and night (not sure about having the sun through a lens onto a CCD device but happy to be proven wrong)
4) can be connected to a bit of computing power (e.g. Raspberry Pi or microcontroller) for analysis of the sensor output

I have an engineering background (although I'm a project manager these days) and I'm not afraid of building things (mechanical or electrical/electronic)

Any ideas?

Susan

Hi

There are a couple of things you could try

My location I made myself a sky can using this software that takes a 360° view of the sky captures the image that I upload this to my website
this gives me a record of the sky conditions through the daytime and there is also a segment timeline view when analysing it gives you an understanding of the cloud conditions and rain during the day

Night time I use a home-made infrared sky temperature sensor this is pretty accurate and gives me a good indication of what happening from clear sky partly cloudy some clouds and totally overcast you can see the data in a graph

The infrared temperature sensor is pointing up at the sky above my location this even though the angle of view is small it does give you an overall sky conditions as the wind moves the clouds over different directions sensor can give an indication of the current sky condition and also and you could use same data for daytime which will correspond to the sky can segment image

If you wanted more angle or a scanning of an area you could attach the sky infrared temperature sensor to a so the motor and control by a Arduino module which you could use in a scanning system

Sky can with infrared data and segment image

http://www.weather-above.com/video.html

WebCam video images and static images
http://www.weather-above.com/meteotemplate/plugins/webcam/index.php

How to make a infrared sky temperature sensor
http://www.weather-above.com/howirsky.html

uv and solar /rain/light add-on modules
http://www.weather-above.com/skyirtemp.html

Hope the above information is helpful and useful

mick
www.weather-above.com

weather-above weather station consists of a  1 wire system ,sky ir cloud temperature ,UV,solar home made modules , live data seismic instrument there is  lot of information on modified sensors,  and 1wire projects pleased  sign my guestbook

Mmichael Parry-Thomas
http://www.weather-above.com/wxgraphic/wxgraphic.php

Offline Bushman

  • Forecaster
  • *****
  • Posts: 7549
    • Eagle Bay Weather
Re: DIY automatic total cloud cover measurement
« Reply #2 on: December 17, 2016, 11:31:54 AM »
Assuming you don't want to just grab values from ClearDarkSky or other websites, I think the easiest thing to do would be to grab a camera with a fish-eye lens (there used to be sorta cheap skycams, but I think  the only place that sold them is done) and grab images from the cam and process them with software.  IIRC there is a post here on the forum about this approach.  It is pretty easy to write code to  bin/slect  pixels based on the colour value.
Need low cost IP monitoring?  http://wirelesstag.net/wta.aspx?link=NisJxz6FhUa4V67/cwCRWA or PM me for 50% off Wirelesstags!!

Offline johnd

  • Forecaster
  • *****
  • Posts: 4849
    • www.weatherstations.co.uk
Re: DIY automatic total cloud cover measurement
« Reply #3 on: December 17, 2016, 01:59:13 PM »
I think the easiest thing to do would be to grab a camera with a fish-eye lens...

What happens when the midday sun's image hits the camera sensor? Can they survive that?
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline Michaelpt

  • Member
  • *
  • Posts: 28
    • Weather-Above.com
Re: DIY automatic total cloud cover measurement
« Reply #4 on: December 17, 2016, 02:57:32 PM »
I think the easiest thing to do would be to grab a camera with a fish-eye lens...

What happens when the midday sun's image hits the camera sensor? Can they survive that?

Hi

I used a very cheap camera positioned over a stainless steel dome when the sun is fully out and you get the reflection from the sun
it just picks up in the image as a nice bright sun image, in the camera itself I have a piece of a lens from a sunglasses which goes darker and more bright
light hit the camera lens  , had no problems been using this camera for years and inside I have mounted inside the housing a small circuit board with small wire wound resistors
which are on all the time to keep any moisture and condensation building up on the camera lens, I even keep it on all through summer and have had no effects to the quality of the images considering I'm using a very low cost outdoor camera, I have tried a fisheye lens on my camera but I was not happy with the results, so I've stuck with the garden stainless steel ball to reflect the sky image into the camera it only good for daytime, I think my camera and cost about £19 .00   
MICK
weather-above weather station consists of a  1 wire system ,sky ir cloud temperature ,UV,solar home made modules , live data seismic instrument there is  lot of information on modified sensors,  and 1wire projects pleased  sign my guestbook

Mmichael Parry-Thomas
http://www.weather-above.com/wxgraphic/wxgraphic.php

Offline Bushman

  • Forecaster
  • *****
  • Posts: 7549
    • Eagle Bay Weather
Re: DIY automatic total cloud cover measurement
« Reply #5 on: December 17, 2016, 04:14:26 PM »
Alcor makes several that handle sunlight.  Although they do say some saturate depending on the model.  Or here's a roll-yer-own option:  http://www.instructables.com/id/Wireless-All-Sky-Camera/  Since it is running on an RPi it would very easy to  process the image for cloud cover (ie. Imagemagick on Raspbian)
Need low cost IP monitoring?  http://wirelesstag.net/wta.aspx?link=NisJxz6FhUa4V67/cwCRWA or PM me for 50% off Wirelesstags!!

Offline graculus

  • Senior Member
  • **
  • Posts: 92
Re: DIY automatic total cloud cover measurement
« Reply #6 on: December 17, 2016, 04:35:30 PM »
Alcor makes several that handle sunlight.  Although they do say some saturate depending on the model.  Or here's a roll-yer-own option:  http://www.instructables.com/id/Wireless-All-Sky-Camera/  Since it is running on an RPi it would very easy to  process the image for cloud cover (ie. Imagemagick on Raspbian)

Quite a high end project, about $400 for that camera and lens.

Offline vreihen

  • El Niño chaser
  • Forecaster
  • *****
  • Posts: 1216
  • K2BIG
Re: DIY automatic total cloud cover measurement
« Reply #7 on: December 18, 2016, 06:35:33 AM »
(Hope this is the right forum - apologies if not but I couldn't see a better one.)

I personally would have posted this to the Weather Sensors and Home Brew child boards under Tech Corner, where it seems right at home:

http://www.wxforum.net/index.php?board=126.0

Sorry that I don't have any input on how to build a cloud measurement sensor, but I'll be following this topic to see if/how it can be done for a reasonable price.....
WU Gold Stars for everyone! :lol:

Offline SLOweather

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 3456
    • Weatherelement Moline IL

Offline graculus

  • Senior Member
  • **
  • Posts: 92
Re: DIY automatic total cloud cover measurement
« Reply #9 on: December 18, 2016, 11:41:18 AM »
Another one Aurora Cloud Sensor.

DIY project from Michaelpt http://www.weather-above.com/skyirtemp.html uses a similar IR sensor.

Offline Bushman

  • Forecaster
  • *****
  • Posts: 7549
    • Eagle Bay Weather
Re: DIY automatic total cloud cover measurement
« Reply #10 on: December 18, 2016, 11:45:43 AM »
For anyone who is interested in using the picture method of classifying sky vs. cloud, here is an interesting explanation uing  OpenCV (computer vision).  In Python it is a couple dozen lines of code max!  http://www.weheartcv.com/pixel-counting-color-statistics-part-1/
Need low cost IP monitoring?  http://wirelesstag.net/wta.aspx?link=NisJxz6FhUa4V67/cwCRWA or PM me for 50% off Wirelesstags!!

Offline Michaelpt

  • Member
  • *
  • Posts: 28
    • Weather-Above.com
Re: DIY automatic total cloud cover measurement
« Reply #11 on: December 18, 2016, 12:59:06 PM »
For anyone who is interested in using the picture method of classifying sky vs. cloud, here is an interesting explanation uing  OpenCV (computer vision).  In Python it is a couple dozen lines of code max!  http://www.weheartcv.com/pixel-counting-color-statistics-part-1/

This is very interesting, I just wished I could get to grips with code as my dyslexia always gets the better of me, my infrared sky temperature project which uses a Python script was written by Niko of weather-watch"

Reading between the lines it looks pretty straight forward to using openCV has anybody else tries putting together automated script taking a image say every 10 minutes
from a web cam image to work out the percentage of cloud cover?
michaelpt
weather-above weather station consists of a  1 wire system ,sky ir cloud temperature ,UV,solar home made modules , live data seismic instrument there is  lot of information on modified sensors,  and 1wire projects pleased  sign my guestbook

Mmichael Parry-Thomas
http://www.weather-above.com/wxgraphic/wxgraphic.php

Offline Michaelpt

  • Member
  • *
  • Posts: 28
    • Weather-Above.com
Re: DIY automatic total cloud cover measurement
« Reply #12 on: December 18, 2016, 01:03:59 PM »
Here's a commercial device that might be good for ideas:

http://diffractionlimited.com/product/boltwood-cloud-sensor-ii/



Here's the manual:

http://diffractionlimited.com/wp-content/uploads/2016/04/Cloud-SensorII-Users-Manual.pdf

The gold bar sticking out from the top side is a wind sensor these are very crude, they use a heated element than the temperature differential is then converted by code
to give a wind speed (the wind speed sensor are very cheap I think they cost approximately £4.00 will have a go at playing with samples I have in the New Year
maybe put a very simple add-on project together

mick
weather-above weather station consists of a  1 wire system ,sky ir cloud temperature ,UV,solar home made modules , live data seismic instrument there is  lot of information on modified sensors,  and 1wire projects pleased  sign my guestbook

Mmichael Parry-Thomas
http://www.weather-above.com/wxgraphic/wxgraphic.php

Offline Aussie Susan

  • Senior Member
  • **
  • Posts: 76
Re: DIY automatic total cloud cover measurement
« Reply #13 on: December 18, 2016, 11:40:46 PM »
Thanks for the suggestions everyone.

With the Christmas break coming up (for us antipodeans it can be a somewhat longer break!) I'll be trying out some of these ideas (like the silver ball refector - I like it!)

Susan

 

anything